image-studio

Pass

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute local Python scripts (generate.py) located at hardcoded absolute paths (C:\Users\renat\skills\...) to perform image generation, editing, and upscaling tasks.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input (image descriptions) and interpolates it directly into shell commands as arguments for the generate.py script.
  • Ingestion points: User input string for image generation requests (e.g., "crie uma imagem de X").
  • Boundary markers: None. User input is placed directly into script arguments without delimiters or guardrails.
  • Capability inventory: Execution of local Python scripts via shell commands (SKILL.md).
  • Sanitization: No sanitization or escaping of the user-provided prompt string is mentioned before it is passed to the command line.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 16, 2026, 09:23 PM